Category archives: Snaking

Looks like two of my talk proposals for PyCon have been accepted.  One is an intermediate/advanced talk porting your code to OpenGL 3.x, the other an introductory presentation on Profiling/optimization.  Should be fun :) .

Back when I was doing the next-generation (now many generations old) Multi-User Technology (MUTech) for Holodesk I thought it was really cool that there was no "server", it was all just peers where any given object could be hosted on any machine, and the system would let you set up (potentially one-way) proxies on any ...

Here are my rough notes from PyGTA's programming mantra's discussion this evening:

In the face of ambiguity, refuse the temptation to guess.
   Have a vision before you start coding.
   Know your task.
   Know your client.
   Ignore the impossibility behind the curtain.
Your clients don't know what they want.
   Charge by the hour.
   Play around with ...

Ran into this problem on Friday at a client site and just couldn't find a Python library that did everything I needed wrt providing login to TurboGears and properly uploading a file to a controller in the TurboGears application.  Bothered me that there was no easy way to do this (seems like it should be ...

Was looking into mitigating some of the slowdown from the recent fix, mainly be rewriting this little function:

@classmethod
def from_param( cls, instance, typeCode=None ):
    try:
        pointer = cls.dataPointer( instance )
    except TypeError, err:
        array = cls.asArray( instance, typeCode )
        pp = ctypes.c_void_p( cls.dataPointer( array ) )
        pp._temporary_array_ = (array,)
        return pp
    else:
        return ctypes.c_void_p( ...

Having to work with svn again at a client site... I miss bzr.  Even bzr-rebase is turning up fondly in my memories.  It's really just the merging... merging with svn seems to always be an "Oh good Lord! We have to MERGE!" event, filled with conflicts and crashes halfway through switching that leave 1/2 of ...

Just discovered an annoying regression in the converters for lists/tuples, namely they will core when you have the accelerator module installed.  They don't core without the accelerator, but that seems to just be dumb luck.  Basically when you pass in a list-of-whatevers to PyOpenGL we have to convert it to something C-friendly, so we want ...

There have been so many changes to try to improve performance and packaging that we're going to have to have another beta for PyOpenGL.  Beta 8 is up on SourceForge.  It includes Win32 installers for both the Python code and the (optional) Cython accelerator modules (for Python 2.5).  Probably should split out the accelerator modules ...

Finally sat down and got my ancient Win2K setup updated sufficiently to install Python 2.6 and MingW32 into it.  Have just compiled the OpenGL_accelerate wrapper.pyd with the setup, but as of yet I don't have enough other 2.6-compatible modules available in the environment to actually test it.  Numpy doesn't seem to be available yet.

SimpleParse ...

Have just been testing the new Cython accelerator module for PyOpenGL 3.x on my workstation.  As seen on the laptop, the PyOpenGL-specific stuff just sort of drops out of the hotspot-set.  There's a few little spots with a percent or two showing up, but pretty much everything else is OpenGLContext scenegraph management stuff.  Those tend ...